3 replies to this topic

[pklavinsJ]

We are modifying an existing oil battery/gas processing facility to handle surface production from a gas/oil/water field. Due to changes to the field production, the asset team would like to handle significantly more water production. This would require increased capacity to send water from the inlet separator to the produced water tanks and thus require either a larger level control valve or parallel level control valves. 

 

If we increase the level control valve size, the gas blowby case for the downstream atmospheric produced water tanks increases beyond the capacity of the existing emergency pressure relief valves. To handle that failure mode, we would need to make costly modifications to these existing produced water tanks to increase the relief capacity of these tanks (or install a costly HIPPS/SIL-3 rated system).

 

An alternative option that we are exploring would be to install a parallel level control valve. This would be controlled from a separate level bridle with a different type of level transmitter with a higher setpoint. At what point would you consider these two parallel level control valves independent from each other (for failure mode analysis) and thus keep the blowby gas flow rate limited to just one of the control valves failing open? Do you consider favorable action of the other level controller/level control valve to close if the liquid level drops creating gas blowby?  

 

I'm not confident that this alternative design will survive a HAZOP analysis and I am trying to eliminate this option early in the project.

[deft clay]

Man this sounds like a controls nightmare. Would the two bridles be at the same levels, just the second one has a higher set point? So basically valve-1 kicks on, then if the rate of liquid flow outruns it, valve-2 also kicks on?

 

I am not upstream so forgive me if this next question is dumb: is it permissible to detect high flow rates of gas and simple shut the nozzle off, trapping the gas in the inlet sep? I ask because there are NFPA valves which do just that. I can dig up some brochures if you are interested. The purpose of them is to cut off flow in the event of a downstream loss of containment... they are intended for propane bullet tanks.

[pklavinsJ]

The two control points would need to be be offset otherwise they will fight with each other. The control system would need to be a lead-lag setup. So if the inflow exceeds the capacity of the lead valve, it would be 100% open and then the level would rise until the second valve opens and balances the inlet and outlet flow. There would also be a dead-band with the lead valve at 100% and the lag valve at 0%. I'm fairly confident that it will work, but level control this way is atypical. I also suspect that this setup does not actually prevent a large gas-blowby scenario from occurring (from an API-521 or risk assessment perspective).
 

Detecting the low level or high gas rate plus a single ESDV would get us one layer of protection. We would need multiple redundant sensors and redundant ESDVs (HIPPS) to get the same reliability as a properly sized relief valve. 

Regarding the NFPA valve - that sounds like an excess flow valve (EFV)? I'm not sure the reliability and testing of such a device would be equivalent to that of an emergency relief valve but it might still provide an additional layer of protection that brings the hazard/risk to an acceptable level. 

 

(Edited for clarity)

Edited by pklavinsJ, 19 June 2025 - 03:58 PM.

[deft clay]

Yeah, that's the type of valve. I had to buy a number of them about 3 years ago. They are all mechanical which is nice.

 

Perhaps a sketch of the vessel and a PFD would get some bigger brains in here to help you.